Cell Selection or Handover in Wireless Networks

1. A system, comprising: a memory storing executable components; and a processor, communicatively coupled to the memory, that facilitates execution of the executable components, the executable components comprising: a candidate component configured to determine active device data corresponding to a first set of network devices and a second set of network devices to use for transfer of communications of a mobile device based on a selection of a network device from among the first set of network devices and the second set of network devices; an evaluation component configured to generate an evaluation via a reward function derived from a decision model that accounts for a profile of a network load level to facilitate network routing decisions to perform the transfer of the communications of the mobile device in a handover process; and a network parameter component configured to determine a set of parameters for the reward function comprising a first network load level and a second network load level based on a current capacity that comprises a current number of connections for the first set of network devices and the second set of network devices.

2. The system of claim 1 , wherein the network parameter component is further configured to determine the first network load level and the second network load level based on a network device loading state transition probability determined from a corresponding previous load state and a set of corresponding previous network device actions of the first set of network devices and the second set of network devices.

3. The system of claim 1 , wherein the executable components further comprise: a reference component configure to determine a reference signal for the first set of network devices and the second set of network devices, and generate a determination of whether a reported reference signal satisfies a threshold condition to facilitate another handover process among the first set of network devices and the second set of candidate network devices; and a Markov component configured to determine, based on the evaluation, the network device of the first set of network devices and the second set of network devices to use to route network traffic from the mobile device based on the reward function being determined from the decision model.

4. The system of claim 3 , wherein the reference signal comprises a signal-to-noise ratio and a mobility prediction of the mobile device, and wherein the threshold condition is a function of a minimum receive signal of the first set of network devices and the second set of network devices.

5. The system of claim 1 , wherein the executable components further comprise: a partially observable Markov decision component configured to account for a set of defined uncertainties in the first network load level and the second network load level.

6. The system of claim 1 , wherein the executable components further comprise: a transition probability component configured to generate a network device loading state transition probability based on a set of previous network device load states at a defined time slot and a set of previous network device actions, respectively associated with the first set of network devices and the second set of network devices.

7. The system of claim 1 , wherein the evaluation component is further configured to evaluate a first set of network parameter conditions of the first set of network devices and a second set of network parameter conditions of the second set of network devices by the reward function based on a mobility parameter of the mobile device and at least one of a network device capacity and a frequency of connection transfers.

8. The system of claim 1 , wherein the evaluation component is further configured to evaluate a first set of network parameter conditions of the first set of network devices and a second set of network parameter conditions of the second set of network devices, via the reward function, that respectively comprises a capacity that is a function of a total network bandwidth and respective signal-to-noise ratios for network device-to-mobile device of the first set of network devices and the second set of network devices.

9. The system of claim 1 , wherein the evaluation component is further configured to evaluate a first set of network parameter conditions of the first set of network devices and a second set of network parameter conditions of the second set of network devices, via the reward function, that respectively comprises a connection transfer time that is determined as a function of a penalty of a transfer of a connection of the mobile device that is represented by a timeslot ratio of a timeslot, a total network bandwidth, and respective signal-to-noise ratios for network device-to-mobile device connections.

10. The system of claim 1 , wherein the evaluation component is further configured to evaluate a first set of network parameter conditions of the first set of network devices and a second set of network parameter conditions of the second set of network devices, via the reward function, comprising a mobility of the mobile device and a time of a transfer of a connection that is determined as a function of a simulated time interval of a continuous movement parameter of the mobile device over a change of time in a set of simulated time intervals, a penalty of the transfer of the connection of the mobile device that is represented by a timeslot ratio of a timeslot, a total network bandwidth, and respective signal-to-noise ratios for respective network device-to-mobile device connections.

11. The system of claim 1 , wherein the evaluation component is further configured to evaluate a first set of network parameter conditions of the first set of network devices and a second set of network parameter conditions of the second set of network devices, via the reward function, comprising a non-fixed signal-to-noise ratio that is based on a current reward of a transfer of a connection to the first set of network devices or the second set of network devices and a predicted reward.

12. The system of claim 11 , wherein the current reward is a function of a capacity at a timeslot for the transfer of the connection and the predicted reward is a function of a path loss estimated at a distance for the transfer of the connection at the timeslot.

13. A method comprising: determining, by a device comprising a processor, a reference signal of a set of candidate network devices; generating a determination of whether a reported reference signal receive power satisfies a threshold condition; and in response to the reported reference signal receive power satisfying the threshold condition, determining a set of active device data associated with the set of candidate network devices to route network communications of a mobile device to a network device of the set of candidate network devices, and routing the network communications of the mobile device to the network device based on an evaluation of the set of candidate network devices via a reward function derived from a decision model that accounts for a profile of a network load level to facilitate network routing decisions.

14. The method of claim 13 , wherein the decision model is a function of a predicted network device load and an active network device load, corresponding to the set of candidate network devices of the set of active device data.

15. The method of claim 13 , further comprising: generating a signal-to-noise ratio and a mobility prediction derived from the reference signal of the set of candidate network devices.

16. The method of claim 15 , further comprising: communicating the signal-to-noise ratio and the mobility prediction derived from the reference signal of the set of candidate network devices based on a predefined period of time.

17. The method of claim 15 , further comprising: communicating the signal-to-noise ratio and the mobility prediction derived from the reference signal of the set of candidate network devices in response to a measurement control message.

18. The method of claim 13 , further comprising: generating a profile of an active network load for the set of candidate network devices via a partially observable Markov decision process of the decision model that is based on a previous load state at a time slot and a set of previous actions of the set of candidate network devices.

19. The method of claim 18 , wherein the set of previous actions comprises at least one of exceeding a capacity limit or executing an off-loading event by the set of candidate network devices to transfer a connection of at least one mobile device connection to a different network device.

20. The method of claim 13 , further comprising: generating another determination of whether candidate network devices of the set of candidate network devices have exchanged network device load data.

21. The method of claim 13 , further comprising: generating the decision model as a function of a predicted network device load that is generated in response to the determination indicating that the set of candidate network devices have not exchanged network device load data, wherein the predicted network device load is a function of a probability of a network device loading state based on a previous load state at a time slot and a set of previous actions of the set of candidate network devices.

22. The method of claim 13 , further comprising: determining the reward function based on at least one of a capacity, a time of a network connection transfer, or a mobility of the mobile device, of the set of candidate network devices.

23. A mobile device, comprising: a memory storing executable instructions; and a processor, coupled to the memory, that facilitates execution of the executable instructions to at least: determine a reference signal of a set of candidate network devices; receive a determination of whether a reported reference signal receive power satisfies a threshold condition; and in response to the reported reference signal receive power satisfying the threshold condition, determine active device data comprising candidate network devices for a potential handover to communicate via a set of network devices of a network, generate an evaluation of the candidate network devices of the active device data comprising a network load and a reward function for decreasing a number of handovers to a different network of the candidate network devices, and determine a network device of the candidate network devices of the active device data to route network communications of the network based on a reward determined by the reward function and the network load.

24. The mobile device of claim 23 , wherein the processor further facilitates the execution of the executable instructions to: route the network communications to the network device based on a capacity according to a partially observable Markov cell selection of a hidden Markov process that selects the network device according to the reward and the network load.

25. The mobile device of claim 23 , wherein the processor further facilitates the execution of the executable instructions to: route the network communications to the network device based on a handover time according to a partially observable Markov cell selection.

26. The mobile device of claim 23 , wherein the processor further facilitates the execution of the executable instructions to: route the network communications to the network device based on a current reward at a time slot determined and a predicted reward via the reward function and a movement measure of the mobile device, according to a partially observable Markov cell selection.

27. The mobile device of claim 23 , wherein the processor further facilitates the execution of the executable instructions to: route the network communications to the network device based on a movement measure and a handover time, according to a partially observable Markov cell selection.

28. The mobile device of claim 23 , wherein the network load comprises an observed number of connections at a time slot respectively for the candidate network devices, and a hidden Markov process selects the network device according to the reward and the network load using a likelihood of the observed number of connections from a historical network load level and a set of corresponding previous actions that comprise at least one of exceeding a capacity limit or respective off-loading events for the candidate network devices.